<![CDATA[The fluctuating nature of Industrial Solar or High Speed Diesel (HSD) sales poses a significant challenge for companies, particularly in developing appropriate distribution strategies and stock planning. This situation demands the application of data-driven analytical methods to support more effective decision-making. This study aims to predict Industrial Solar sales at PT Heva Petroleum Energi Palembang using two Machine Learning methods, namely Support Vector Machine (SVM) and Random Forest. The data used are monthly sales records for the period 2022–2024. The research process includes data collection, pre-processing with normalization and feature selection, model building, testing by dividing the data into training and test sets, and performance evaluation using the Mean Absolute Percentage Error (MAPE) metric. The results show that the Random Forest model produces a MAPE value of 12.48%, while the Support Vector Machine model obtains a MAPE value of 12.97%. This comparison shows that Random Forest is superior in predicting sales compared to SVM. Thus, it can be concluded that Random Forest is a more appropriate choice for application in modeling Industrial Solar sales. The implications of these findings are expected to provide a real contribution to companies in developing distribution policies and stock management that are more accurate, efficient, and sustainable, so as to be able to support the stability of company operations in the future.]]>
Copyrights © 2025
